ECONOMIC INVESTIGATIONS OF SHORT ROTATION INTENSIVELY

CULTURED HYBRID POPLARS

David C. Lothne I/

Abstract .--The history of the economic analyses is sum-

marized for short rotation intensively cultured hybrid

i poplar at the North Central Forest Experiment Station.Early break-even analyses with limited data indicated that

i at a price of $25-30 per dry ton for fiber and low to

i medium production costs, several systems looked profitable.! Later cash flow analyses indicated that two systems without

irrigation and fertilization could achieve internal rates

of return of 8 percent with a 5 percent inflation rate.However, two systems with irrigation and fertilization had

negative internal rates of return. All systems had nega-tive net present values when a i0 percent discount rate was

used. We are currently incorporating risk analysis into

our financial investigation to directly account for uncer-tainty in the performance calculations.

During the last several years interest has The economist can aid both. He can aid the

increased remarkably in intensive agronomic manager with financial analysis models that

techniques to produce high yields of woody describe cost and return relations, determinebiomass for energy and fiber. And our capital requirements, and calculate f_nancialknowledge and experience through many studies performances of alternative projects. He can

and field trials has been increasingly aid both manager and scientist by pointing out

encouraging. Land managers and scientists in the importance of the various activities to theboth the public and private sectors are facing overall financial performance. And, finally,

several important decisions, he can suggest the most fruitful areas forfuture research from a financial perspective.

For the land manager few tasks are morechallenging than developing a production plan In this paper I will summarize past inten-

to meet future mill requirements. Options sive culture economic research sponsored by the

include: investing in intensive culture short- North Central Station, discuss present on-

rotation (SRIC) forest plantations, investing going work, and make some general conclusions.in conventional less-intensive longer rotation

forest plantations, or purchasing the wood as

needed on the open market. If a decis'ion is BREAK-EVEN ANALYSISmade to invest in SRIC forest plantations, how

intensively should these plantations be managed When early research efforts began, SRIC

and which of numerous management perscriptions systems were believed to have several advan-should he followed? tages over conventional silviculture systems.

SRIC systems could produce more biomass from

For the scientist it takes time to develop the same acreage or the same quantity of

information about specific management alter- biomass from less acreage. SRIC systems could

natives and the number of possible management also produce this biomass much sooner--5 to 15alternatives is enormous. Scientists must years after establishment rather than the 30 to

decide which aspects--site preparation, 80 years of conventional silviculture.

planting, post-planting release, etc.--of the Finally, the plantations could be regenerated

overall management system merit further study by coppicing every 5 to i0 years from the same

in order to improve performance, root stock, thus eliminating the need forperiodic replanting.

A major disadvantage was that SRIC systems

1/Principal economist, USDA Forest require large capital expenditures for higherService, North Central Forest Experiment quality, accessible land as well as for all the

Station, Duluth, Minnesota

139

machines and installations required to perform Several different production systems were

the intensive management. This is particularly evaluated using this technique. The major

true for irrigated plantations, variables were spacing, rotation length, numberof coppices, cultural practices (e°g.,

Thus, the economic potential of alternative replanting), and project length (table i).

SRIC systems generated much discussion as the Estimating production cost is extremely impor-research program began and break-even analysis tant in a break-even analysis. However, little

was selected to make preliminary economic eval- actual data were available when the early anal-

uations (Rose 1976, Rose and Kallstrom 1976, yses began, complicating the situation. Thus,and Rose 1977). With break even analysis, the each of the production systems was evaluated

compounded cost of growing and harvesting under low, medium, and high cost assumptionsbiomass is calculated under alternative manage- for two different initial site conditions:ment strategies. This in turn is compared to: forest and pasture.

(I) the physical yield (e.g., dry

tons/acre/year) at a specified price, or (2) These early break-even results estimatingthe price (e.g., dollars/dry ton) at a speci- the cost per dry ton of fiber produced are

fled physical yield. The break-even point is shown in table 2. Alternatives 5, 6, 7, 8,the dollar value of the yield that just covers and i0 would be profitable assuming both a

all compounded production costs, price of $25-$30 per ton for dry fiber and pro-duction costs at a low to medium level.

Alternative 3 would be profitable only when thesame prices are assumed and costs are low.

Table l.--Specifications for intensive cultural alternatives

I..........1Alternative Spacing Rotations Stand Project Regeneration

(feet) years) (years) (age)

1 4x4 (I) 4 cuttings 20 none

(4) 4 coppice

2 4x4 (2) 4 cuttings 24 13

(2) 4 coppice

3 4x4 (1) 5 cuttings 20 none

(3) 5 coppice

4 2x4 (2) 2 cuttings 20 ii

(4) 2 coppice

5 12x12 (i) i0 cuttings 20 none

(i) i0 coppiceI

6 12xl2 (2) lO cuttings 20 ll

7 12x12 (I) 15 cuttings 30 none

(i) 15 coppice

8 4x4 (i) 15 cuttings 30 none

(I) 15 coppice

9 4x4 (i) 5 cut ting s 20 none

(3) 5 coppice!iiii

i0 12x12 (I) 15 cuttings 15 none

140

Table 2,o--Break-_-eve<__>esu].tsof ten intensive calculated from harvest yield assumptions based._,(.,_,__ture_Jiternatives on experimental plots that were irrigated and

fertilized annually (Ek and Dawson 1976). In

-A].[i:_f_"°i-_iilT{i{-:_i-"-l_:[![ti7_][°L_-i[ti_i-]-Cost Range/Ton contrast_ most of the alternatives shown inCondition __ of Fiber table 2 assumed no irrigation and only periodic

I Produced fertilization° If the management alternatives(low) (high) were carried out on sites where water andr_'o

],_ Foresu_ $ _)_o48-S.97°2]..... nutrients were lffm_itinF_factors, the costs

Pasture 43°82- 84,71 would be higher, Thus, these first analyses

were crude and only roughly approximated the

2 Forest 54°89- 100o59 economic potentials of alternative production• o7,:, 68°75 . Further investigations wou]d _]?asture 4.5 __- systems

necessary using more sophisticated models when

3 Forest 29°82- 58°74 better production cost and yield informationPasture 24°26-- 51o56 became available°

, q]'o4 For'est .......2 78-739,46

Pastu.se 436o54- 652 °02 CASH FLOW ANALYSIS

5 Fo<:est 21o68- 44oF_8 As research on SRIC progressed, morePs_At_,r'e 17o30- 39o18 realistic information became available on pro-

duction costs and biological, growth and yield,

6 Focest 22°23- 45°58 and land managers wanted more specific economicPasture 17o82- 39°88 info_xnation about short-rotation systems, we

chose to analyze them with a cash flow model

7 Forest 27o11 (Lothner et al° 1981, Rose et alo 1981_ and

Pasture 22°65 Ferguson et alo 1981)o

8 Forest 26°86 We evaluated a few specific hybrid poplar

Pasture 22°67 plantations representing a likely range of

spacings (4 by 4 feet and 8 by 8 feet), rota-9 ............ 35°34- 60°64 tions (5, I0 and 1.5years), and cultural prac-

I0 18o23- 42°08 tices (including irrigation and fertilization)_Four alternative systems combined these vari-

Production cost pei: dry ton differs ables (table 3). Each system consisted of 1,000

greatly dependir_g upon the general cost level acres of cleared marginal agricultural ],andassumed° Also_ these production costs were arranged in I0 tracts of 80 to 120 acres each.

7_'a_,__=__e_.3°--.Four alternative SRIC systems

............................................................Rotations Dry Yields

Length I Stand Origin Per Acre/[ Total Yield atAlternatives Year End of Rotation

........................................................................................(years) (tons) (tons)

4 by 4 foot spacing I0 cuttin._s 6°3 63

irrigated and 5 coppice 7,2 36fert i!ized 5 coppice 7,2 36

5 coppi ce 7,2 36

5 coppice 7,2 36

4. by 4 foot spacing I0 cuttings 3.2 32not irrigated or 5 coppice 3,6 18fertilized 5 coppice 3,6 18

5 coppice 3.6 18

5 coppice 3,6 18

8 by 8 foot spacing ].5 cuttings 6°3 95

irrigated and 15 coppice 6.3. 95fertilized

8 by S foot spacing 15 cuttings 3_2 47not irrigated or 1.5 coppice 3.2 47

r"+"4 ] ,{,_ •fe _.<...,_......zed

141.

The important production and management Some of the cost and return assumptions

activities of each system were identified by follow:North Central scientists and the forester in

charge of the only large-scale industrial SRIC Cost Estimate (1979 dollars)

plantation in the Lake States. Physical anddollar estimates were made for each of the pro- Land $400 per acre

duction factors and yields as well as the rela- Planting stock $ 80 per thousand

tive uncertainty of each. Using a cash flow Fuel (diesel) $ i per gallon

model, the financial performance of each system Labor (nonunion) $ 4 to $ 6 per hour

was evaluated by an internal rate of return Site preparation(IRR) and present net value (PNV) criteria for & weed control $33.42 to $73.11 per acre

a 30-year period. We also evaluated the sen- Irrigation $99.74 per acre per year

sitivity of the IRR and the PNV to both rela- Irrigation equip- $387,700(initial investmenttive and absolute changes in the production ment for 1,000 acres)

factors and yields due to uncertainties. HarvestWhole tree $ 14 to $ 18 per dry ton

Site preparation for the plantations Forage $ 8 per dry ton

included plowing, disklng, and applying Administration $7,500 per year (for I000

preplant herbicides. All these activities took acres)

place in late summer and fall prior to spring Property tax $ 4 per acre per yearplanting. Three cultivations and two applica- Income tax 28 percent of capital gainstions of herbicide were used for postplanting Equipment 1 percent of new price per

weed control. Trees were irrigated by a tray- insurance year

eling gun system that applied I0 inches of_f _ water per acre annually. Liquid nitrogen fer- Return

__ tilizer was applied in the irrigation water at Product value $25 per dry ton_ I00 pounds per acre per year. Plantations were

harvested using whole-tree chippers for the I0- A more detailed account of the costs andand 15-year rotations, and forage harvesters returns is found in Lothner et al. (1981) and

/ for the 5-year coppice rotations. Expenditures Rose et al. (1981).

common to all alternatives included administra-

tive costs, insurance, land purchase and sale, With the assumed I0 percent discount rate,

equipment cost, and taxes (income and no alternative had a positive PNV. However,

property), the two systems without irrigation and ferti-lization had positive after tax internal

An annual inflation rate of 5 percent and rates of return of 8 percent; the two irrigateda I0 percent discount rate were applied to all and fertilized systems had negative internal

costs and returns, rates of return (table 4). Economically, dif-

Table 4.--Initial investment and investment performance of four SRIC alternatives

Management Alternative Initial ii Present Net I Internal RateInvestment | Value | of Return

(dollar/acre_ (dollar/acre_ (percent)!_!_._, 4 by 4 foot spacing, irrigated 1,158 -2003.82 -0.4

_i_ and fertilized, short-

i_7_ rotations (5 to I0 years)

4 by 4 foot spacing, not irrigated 770 - 236.78 8.1or fertilized, short-

rotations (5 to I0 years)

8 by 8 foot spacing, irrigated 945 -2149.51 -1.6

and fertilized, long-

rotations (15 years)

8 by 8 foot spacing, not irrigated 557 - 200.30 8.1or fertilized, long-

rotations (15 years)

i Includes land purchases, initial equipment purchase or lease, site prepara-tion and planting costs, and administration cost for first two years.

.....

142

ferences were small between rotation length and next important factor. For the 8 by 8 foot

spacing options, but they were significant spacing alternative, a I0 percent change in har-between the systems that were irrigated and vesting and transport cost changed PNV by almostfertilized and those that were not. 25 percent or about $49 (fig. I).

A sensitivity analysis can help identify For the irrigated and fertilized systems, irri-how each cost or revenue factor affects the gation operating costs, primarily fuel, was the

financial performance. In all four systems, second most important factor affecting invest-

the most important factor affecting investment ment performance and would have to be substan-

performance was the product sale value. It was tially reduced for these systems to becomealso one of the most uncertain factors. A financially attractive. Fuel costs represent 68

change in yield, product price, or both, could percent of the operating costs of the travelingsubstantially change the economic attrac- gun irrigation system and a substantial fuel

tiveness of SRIC systems. For example_ a I0 cost reduction with this type of irrigation

percent increase in mill prices from $25 to system seems unlikely. However, new technology$27.50 per dry ton for chips for the 8 by 8 or changes in the irrigation schedule to only

foot nonirrigated and nonfertilized system sporadic use (e.g., during establishment) couldincreased the PNV by 44 percent or $88 (fig. significantly reduce irrigation operating costs.

I).Site preparation costs and planting costs

For systems without irrigation and fertili- for any of the systems have little effect onzation, harvest and transport cost were the PNV. For example, a I0 percent change in site

Figure lo--Sensitivity of net present values (NPV) for 8 by 8 foot spacing, nonirrigated andnonfertilized system to relative changes in costs or prices.

+60

/W PRODUCT PRK_E+40 iO

¢*20-

JO

PLANTING COST>

z_20.z StaTEPREPCOST

u,I_-40' LANDCOSTZ¢_ HARVEST &_

TRANSPORT _O -60. COST

-20 -lo 0 +1o .2oCHANGE iN COST OR PRICE (PERCENT)

1'43

:ii_i

preparation cost for the 8 by 8 foot, nonirri- We are currently cooperating with Michigangated and nonfertilized system changes PNV by 4 State University-- to model the risk inherent in

percent or $8, while a I0 percent change in SRIC projects. This work is based on some

planting cost changes PNV by 3 percent or $6 earlier work in risk analysis by Hertz (1964)

(fig. i). and developed and reported on by others(Davidson and Cooper 1980; Economos 1968; and

The cash flow analyses suggested that Hillier 1963). Incorporating risk analysis into

nonirrigated and nonfertilized SRIC systems a basic cash flow model requires three basicwould be preferable to those that are irrigated steps. First, a range of values is estimatedand fertilized. But because the returns on for each of the production activities (eog_, a

investment would be low even for the nonirri- range of site preparation costs) and the proba-

gated and nonfertilized systems, these systems bility or likelihood of occurrence of different

would be of interest only to individual users values within the range. This is in contrast to

of wood biomass who can compare the potential the previous practice of estimating only theof these systems to alternative sources of most likely value of each factor. Second, one

supply prior to making an investment decision, value from the distribution of values for each

factor is selected at random. The resulting setIn addition, the results indicate that the of selected values for all factors are combined

investment performance of the SRIC system is in one cash flow analysis and the financial per-

most sensitive to the product sale value (yield formance (e.g., internal rate of return or pre-

and product price), irrigation operating costs, sent net value) is calculated. Third, the

and harvesting costs. Our current data con- second step is repeated over and over again tocerning all these factors are inadequate, develop a probability distribution for the

However, by pointing out their importance, we financial performance outcomes. This distribu-

can focus our research efforts on them in the tion of financial performance outcomes thus

future. This may help us to reduce some of the shows the probability of achieving any specifiedimportant sources of uncertainty and make more outcome--e.g., internal rate of return.

informed future decisions. However, many ofthe factors depend on uncertain future events If we use two of the hypothetical SRIC

for which there will always be a limited pre- systems mentioned earlier--the 4 by 4 foot and 8sent understanding° by 8 foot nonirrigated and nonfertilized

ii

systems--we can illustrate what a risk model can

However, some additional risks were not provide in comparison to the conventional cashaddressed by the cash flow analysis. For flow model. Let me emphasize that the numbers

example, what are the risks of insect or found in this section using the risk analysis

disease outbreaks and what impact would they model have no empirical value. They have been

.... have on the investment performance? Or, what selected only to illustrate the risk analysisis the risk of failure, in unirrigated and method°

unfertilized plantations _

............. Using a conventional cash flow model, bothRISK ANALYSIS the 4 by 4 foot and 8 by 8 foot nonirrigated and

• $inonfertilized alternatives show 8 percent inter-

In a conventional cash flow analysis, hal rates of return. Incorporating risk esti-Jj_ mathematical formulas are used to precisely mates provides the decisionmaker with estimates

calculate financial performance estimates for of the probabilities of achieving each possiblevarious criteria by incorporating point esti- value for the financial performance criteria--mates for each relevant factor included in the e.g., internal rate of return. Let's assume

analysis. However, both the analyst and the that when we incorporate risk into our analysis,

decisionmaker--manager or scientist--know that the most likely outcomes for both the 4 by 4

behind these precise calculations are imprecis_/ foot and 8 by 8 foot alternatives remain the

data. A great deal of uncertainty and/or riskZ-" same as the conventional outcome--an 8 percentalways surrounds the basic information that goes internal rate of return (fig. 2)_ In practiceinto the analysis. Our past cash flow calcula- the most likely internal rates of return foundJ

tions have not adequately measured risks nor using a risk analysis model would probably dif-

have we been able to directly incorporate risks fer from the expected internal rates of return

into investment performance calculations, found under the conventional method using singleiii!!!i

point estimates. In the example above, we have

2/ In our discussions a risk situation is where

the probability of an activity or event --3/Paul A. Rubin, Assistant Professor, Graduate

occurring is known or can be estimated;anprobabili_ StateSCh°°lOfuniversity.BUsinessAdministration, Michiganuncertainty situation is where theties of an activity or event occurring have not

been estimated or determined.

144

Figure 2,--Probability distributions of the internal rate of return for two hypothetical SRICsys terns

assumed both alternatives have about an 8 Internal Rate Probability of achieving at leastpercent chance of achieving the 8 percent of Return the returns indicatedinternal rate of return. The chances or proba- Percent Percent

bility of a zero or negative internal rate of 8x8' spacing 4x4' spacingreturn (no profit at all or even a loss) are

much greater for the 4 by 4 foot system (fig. 0 99 94

2). (Remember that these numbers are for 2.5 96 83

illustration purpose only and have no empirical 5.0 87 65value.) 7.5 69 45

8 .I 66 42Z

Using a risk analysis method, the uncer- I0o0 50 25

tainty of achieving at least a certain finan- 12.5 29 8

cial performance--i.eo, internal rate of 15.0 14 0return--can be clearly displayed: 17,5 3 0

2O°0 0 0

We can compare the cumulative profile for

the 8 by 8 foot and the 4 by 4 foot spacing

systems using a risk analysis with the profilefor these systems obtained under the conven-

tional approach (fig. 3). Under the conven-

145

Figure 3.--Cumulatlve rate of return profile for two hypothetical SRIC systems using a conventionalcash flow model and a risk analysis model.

0=w 100 "-"."_ _'_""......... "'":_=m o 0 0 #" .CONVENTIAL ANALYSIS,

_ o 00

W " o "4 X 4' AND 8 X 8' SYSTEMSIX: • .

800 •

eoe

0 * "

_; "J RISK ANALYSIS, • _:_i__ _ ---D P- 60 :o

_ 0 Z 4 g 4 SYSTEM •

V "W ig . : o

(f) : RISK ANALYSIS,

== _ 40_: --_ . 8 X 8' SYSTEM=mm : • 0

t

• 0>-_,- * °mmm •

-J 20 . o• 0

• 0

in .O 41 I O

®• O

-5 0 +5 10 15 20

iNTERNAL RATE OF RETURN (%)tlonal cash flow approach we would have no way to the analyst for making clear choices. Anto choose between the two systems because both analysis between two alternatives could result

produced an 8 percent internal rate of return, in one alternative being both more risky as

However, using risk analysis we can see that well as having a greater opportunity for largerthe 8 by 8 foot spacing would be superior to gains (fig. 4)°the 4 by 4 foot spacing. With the wider

spacing we are 99 percent certain that we will SUMMARY AND CONCLUSIONSnot obtain a negative internal rate of return

as compared to only a 94 percent chance with Economic analysis of SRIC at the North

the narrower spacing alternative. In addition, Central Forest Experiment Station has

we have a 66 percent probability of achieving progressed during the past several years.

at least an 8 percent rate of return with the After some early break-even analyses when basicwider spacing system as opposed to only a 42 data required for economic evaluation werepercent probability with the narrower spacing, almost nonexistent, we moved on to discounted

It is clear In this example that the wider cash flow analyses using sensitivity analysisspacing system is significantly better, to handle risks and uncertainties. Through

Without accounting for risk in the model as these analyses we learned that using single

with the conventional cash flow model, we would estimates of production factor values, the

never know this. monetary returns were fairly low. Through thesensitivity analysis we learned that financial

This is not to say, however, that using performance was most sensitive to product

risk analysis will always present information price, product yields, irrigation costs, har-

146

Figure 4.--Alternative probability distributions of the internal rate of return for twohypothetical SRIC systems.

I=. _ 2 X 8' SYSTEMZ

0

UJ

4 X 4 SYSTEM

0_ .

, _ "II

-5 O *5 10 15 20INTERNAL RATE OF RETURN (%)

vesting costs_ and land costs. These are all Much work remains. If we recognize uncer-

highly uncertain factors. The traditional and tainty, how do we build the model to allowmore certain management activities necessary to managers to react to possible outcomes? And,

establish and grow hybrid poplars, such as site once the model is developed, are there any real

preparation and planting, were not as important advantages for evaluating SRIC investment

as the previous factors° We also believe that opportunities by this method? These will takeconsiderable risk is involved in analyzing the considerable effort. Yet, the discipline of

financial performance of these systems where thinking through the uncertainties of the

uncertainties are not directly accounted for problem will perhaps in itself help the analyst

within the system° assist the decisionmaker in making wise invest-ment choices.

From a growing awareness of the importance

of uncertainties we began to develop models to

handle this problem. Work is still in progressso we do not have any real results. However,

we can use hypothetical examples to show the

kind of information that can be obtained by

incorporating risk directly into the financial

analysis.

147

/

iI

i!

LITERATURE CITED

Davidson, Lynn B., and Dale O. Cooper. 1980. Lothner, David C_, Dietmar Rose, J.

Implementing effective risk analysis at Zavitkovski, and Karen Ferguson. 1981.Getty Oil Co. Interfaces 1(6): 62-75. A financial analysis of poplar intensive

cultures in the Lake States. In Energy

Economos, A.M. 1968. A financial simulation from Biomass and Waste V: Symposiumfor risk analysis of a proposed subsidiary° papers, p. 145-163. Institute of Gas

i

Management Science 15(12): 675-682. Technology. Lake Buena Vista, Florida.

Ek, A. R., and David H. Dawson. 1976. Yields Rose, Dietmar. 1976. Economic investigationsof intensively grown Populus- actual and of intensive silvicultural systems. Iowa

projected. In U.S. Dept. of Agriculture State Journal of Research. 50(3):Forest Service General Technical Report 301-315.

NC-21, p. 5-9. U.S.D.A. For. Serv. North

Central For. Exp. Sta., St. Paul, MN. Rose, Dietmar. 1977. Cost of producing energyfrom wood in intensive cultures. Journal

Ferguson, Karen, Dietmar Rose, David Co of Environmental Management. 5: 23-35.Lothner, and J. Zavitokovski. 1981.

Hybrid poplar plantations in the Lake Rose, Dietmar, Karen Ferguson, David C.Lothner,............. States - a financial analysis. Journal of and J. Zavitkovski. 1981. An economic

_ Forestry. 79(1): 664-667. and energy analysis of poplar intensivei cultures in the Lake States. U.S.

Hertz, David B. 1964. Risk analysis in capi- Department of Agriculture Forest Service,

_ _ ...._ tal investment. Harvard Business Review. Research Paper NC-196, 44 p. U.S.

42(1): 95-106. Department of Agriculture Forest Service,North Central For. Exp. Stao, St. Paul, MN

Hillier, Frederick S. 1963. The derivation of

..... probalistic information for the evaluation Rose, Dietmar, and Robert D. Kallstrom. 1976o

of risky investments. Management Science Economic feasibility of intensive

9/(3): 443-457. cultures. In U.S. Dept. of AgricultureForest Service General Technical Report

NC-21, p. 96-107. U.S. Dept. of

Agriculture Forest Service. North Central

Forest Experiment Station, St. Paul,Minnesota.

!i!! )

148